RE: Programming a trigger audit in ILE

[Buck Calabro <Buck.Calabro@xxxxxxxxxxxx>]

>Buck, null values in SQL are short integers, but in the trigger buffer
>they are single-byte characters (RPG indicators).
>
>> DRtnNullRaw       ds                  occurs(10)
>> D NullMap                         n   dim(9)

Barbara's absolutely right.  I pulled the example from the wrong program.
My trigger maps look like this:

D TrgBuf        E DS                  EXTNAME(Trigger)          
                                                                
 * We can't tell how many fields are in the file;               
 * to avoid hard-coding the number, set up an arbitrarily large 
 * array to hold the null map.  Remember that you can only      
 * trust the array up to the null map length!!!                 
D OldNulImg       ds         32766    Based(POldNul)            
D  OldNulMap                     1a   dim(%size(OldNulImg))     
D NewNulImg       ds         32766    Based(PNewNul)            
D  NewNulMap                     1a   dim(%size(NewNulImg))     

 * Map the file I/O buffers from the external definitions
D OldRcdImg     E DS                  EXTNAME(DATESAMPLE)
D                                     based(pOldRcd)     
D                                     prefix(O_)         
D NewRcdImg     E DS                  EXTNAME(DATESAMPLE)
D                                     based(pNewRcd)     
D                                     prefix(N_)         

C     *Entry        Plist                            
C                   Parm                    TrgBuf   
C                   Parm                    TrgBufLen
C                   Parm                    TrgRtnCde

C                   EVAL      pOldRcd = %ADDR(TrgBuf)+ Z1OROFF
C                   EVAL      pOldNul = %ADDR(TrgBuf)+ Z1ORNBO
C                   EVAL      pNewRcd = %ADDR(TrgBuf)+ Z1NROFF
C                   EVAL      pNewNul = %ADDR(TrgBuf)+ Z1NRNBO

Sorry for the snafu, and thanks, Barbara!
  --buck